Hospital system catches 70% of lung cancer cases in earliest stage, thanks to AI intervention
With the help of artificial intelligence support and an electronic health record tracker, one hospital system has significantly improved the diagnosis of early-stage lung cancer.
California-based Sutter Health operates dozens of acute care hospitals and more than 200 clinics throughout the state. As such, the system conducts thousands of routine imaging exams daily. These exams sometimes include incidental findings that can be overlooked by radiologists, allowing potentially problematic diseases to go unchecked and untreated for years.
Some of the most common incidental findings on CT scans are pulmonary nodules. While not all nodules are indicative of lung cancer, some represent early-stage malignancy. Lung cancer is the leading cause of cancer death in the United States and globally, but when caught early, patient outcomes improve significantly.
This is where the identification and communication of incidental pulmonary nodules comes into play. Often, even when radiologists comment on their presence in radiology reports for exams completed for other clinical concerns, a lack of exclusive communication prevents patients from following up on the findings.
To address this, Sutter Health recently implemented an AI solution into its CT workflow that flags pulmonary nodules—incidental or otherwise—that are 5 mm and smaller. This alerts radiologists to their presence; after receiving the alert, readers leave a remark pertaining to the nodule in the patient’s report and within the patient’s chart via Epic’s Study Review. The EHR notification generates an alert in a shared and monitored tracking list indicating the patient needs to follow up.
This prompts nurse navigators initiate communication between providers, triggering a review process to determine the best follow-up procedures for the patient; providers move forward with the next steps accordingly. The closed loop communication system has significantly improved the early diagnosis of lung cancer for Sutter Health patients. Now, more than 70% of lung cancer diagnoses within the health system are made while still in stage 1 or 2.
“It takes my breath away to think about how we inverted that curve—from mostly late-stage diagnoses to now catching over 70% of cases early, when survival is vastly improved and many patients can be treated with curative intent,” Jason Wiesner, MD, chief radiologist and director of imaging informatics at Sutter Health, said in a recent interview with American Medical Association.
The system was developed to be adaptable within the EHR, enabling improved communication of incidental findings beyond just pulmonary nodules to include other high-risk findings, such as brain aneurysms.
“The EHR is the backbone of this program,” Wiesner said. “Everyone working in the same instance, with a single medical record across the health system, allows for seamless handoffs.”
Although Wiesner acknowledges organizations’ concerns related to the costs of implementing such systems, he says the investment is worth the improved patient outcomes.
Read his full interview with AMA here.
In addition to her background in journalism, Hannah also has patient-facing experience in clinical settings, having spent more than 12 years working as a registered rad tech. She began covering the medical imaging industry for Innovate Healthcare in 2021.